TPS voltage drifting on roller barrels

[KnifeySpoony]

I've been dealing with a part throttle/low load hesitation/misfire on my 420R (w/roller barrels) - https://www.caterhamlotus7.club/forum/420r-hesitation-part-throttle-when-hot

 

I thought I had it figured out after realizing that my TPS voltage had drifted up. It's supposed to be 4.62V at idle. I found that it was 4.71V and this seemed to be the cause of my hesitation issue. Resetting it to 4.62 quickly allowed the engine rev normally. However I saw that the voltage was hunting a bit just sitting there connected to easimap, so I decided to replace it. The first sensor I put in I set to 4.62, then did a short drive. After the car got warm (about 5-10min) I noticed the idle speed had crept up to 1800rpm (normally ~1100rpm). Sure enough when I got back, the voltage was 4.55 (too low a voltage causes the ECU to think more throttle is applied as the voltage from the TPS goes down as the throttle is opened). So I thought maybe that sensor was bad. I got a new OEM Ford TPS sensor and swapped it in yesterday, set to 4.62 of course. Took the car out again and noticed that it was still doing the part throttle hesitation, although just a tiny bit and barely noticeable. Totally driveable, not the severe misfire it had before, and only over a tiny range of the throttle sweep (ie dead spot between 10-12% throttle, where before it was severely sputtering between 1 and 40%). However the idle speed had gradually crept up to 2000rpm. I got home (with the car still hot) and checked the voltage, which was now 4.51V. Needless to say I am confused. The sensor is fastened tightly - I don't think it's moving. Could this be a temperature issue? The TPS does get quite hot being mounted directly to the roller barrel assembly. I don't think Ford intended this part to get this hot. Could the lower voltage be caused by increased resistance in the wiring with heat? I have jiggled the wiring while connected to easimap and didn't see the voltage change at all.

 

Any thoughts are appreciated.

 

 

[KnifeySpoony]

BTW I reset voltage when I got home with the car hot. Will check in AM with car cool to see if it changes again.

[revilla]

It's probably worth checking that the 5V reference voltage supplied by the ECU (measured across the two end terminals of the TPS) isn't drifting. The 4.62V you are setting is just a fixed fraction of the 5V supply so will drift with it.

The ECU will probably use that same 5V as the analog reference when digitising the TPS voltage so any small changes should come out in the wash, but it really should be accurately regulated so if it does drift by much it means there's a problem somewhere.

[KnifeySpoony]

Checked voltage this AM after car cooled. Voltage still at proper setting after adjusting yesterday afternoon hot. Went for shortish drive, came back and voltage stable with car hot. I don't get it.

 

Line voltage at the plug was 4.92V - which was on 2 out of the 3 pins (not sure I understand that). Not sure if that being under 5V is significant or not.

[revilla]

With the TPS plugged in and connected, one end terminal should read ver close to 0.00V (check the voltage between the terminal and a good earth point). The other end should read close to 5.00V. The middle terminal voltage should rise and fall as the throttle moves.

It's probably not so critical that the 4.92V isn't exactly 5.00V, it's more important that it doesn't change when the engine gets hot / when the ECU gets hot / when things get shaken around on a drive. It needs to a steady, reliable, constant voltage.

For example, if the supply to the TPS is reading 5.00V and you adjust things to get 4.62 exactly, if the supply voltage then drifts down to 4.92V then straight away the TPS voltage will fall to 4.92 / 5.00 * 4.62 = 4.55V.

If the "5V" supply isn't accurately regulated you'll be chasing your tail trying to get a fixed TPS voltage even if nothing else is changing.

[Markiebabes]

Is the TPS a spring loaded one?

I found it seemed to flicker at idle where as the non-spring loaded one seemed fine!

Worth checking. 

 

[KnifeySpoony]

Not spring loaded. Just a freely turning pot.